A Prominent Climate Skeptic Now Affirms that Global Warming is Real

Summary.  Richard A. Muller, professor of physics at the University of California, Berkeley, has long been a skeptic concerning the validity of the temperature trends that characterize recent global warming.  His doubts related to the accuracy and statistical significance of the temperature records used, and to the methods of statistical analysis employed, in support of global warming.  Now, he and a team of researchers have reanalyzed the data, extended the size of the database of temperature records, and developed improved statistical methods to analyze the data.  Their results confirm that  global warming is real, having warmed by just under 1.0ºC (1.8ºF) since the mid 1950’s.

Introduction.  Richard A. Muller, a respected professor of physics, but not a climate scientist, at the University of California, Berkeley has been a long-standing skeptic concerning the reality of global warming.  Paul Krugman wrote in the New York Times  that he criticized former U. S. Vice President Al Gore and others who warned of the hazards of continued global warming as being “exaggerators” of the phenomenon, and that he was active in the 2009 controversy over emails from climate scientists at the University of East Anglia (“Climategate”).  The Sydney, Australia Morning Herald writes that Muller used to dismiss alarming climate research as “polluted by political and activist frenzy”. Don Shelby, on Minnpost.com, writes that Muller has long been critical, in particular, about the temperature measurements that climate scientists around the world use in assessing global warming.

Prof. Muller has now confirmed that global warming is occurring.  He has recanted his former skepticism concerning the validity of the rise in global temperature.  Over the past two years he directed the Berkeley Earth Surface Temperature Project (BEST) whose objective was to reexamine all temperature records available for two hundred years in order to arrive at an authoritative assessment of temperature changes.  BEST has received important funding from the Lee and Juliet Folger Fund ($20,000), the William K. Bowes, Jr. Foundation ($100,000), the Fund for Innovative Climate and Energy Research (created by Bill Gates) ($100,000), the Charles G. Koch Charitable Foundation ($150,000), and the Ann & Gordon Getty Foundation ($50,000).  The Koch family, for example, with business interests in fossil fuels, supported the failed attempt in California to overturn by referendum the state’s ambitious program to reduce greenhouse gas emissions.  BEST has posted its methods and results online, in four manuscripts, simultaneously with having submitted the papers for peer review by a scientific journal. At the same time, Muller described the BEST project in a recent op-ed in the Wall Street Journal.

Summary of Muller’s doubts about temperature measurements.  In the Wall Street Journal, on the web site, and in the written papers Muller identified specific problems in earlier analyses by other groups of the temperature record that have concerned him.  First, as many as 757 weather stations in the U. S. recorded a cooling over the past century, many of which are in the southeast where tornadoes and hurricanes are prevalent.   Second, because of poor siting of stations in the U. S., many have reported temperatures that may be in error by as much as 2-5ºC (3.6-8.0ºF), according to the U. S. government.  He surmises that there may be many more stations reporting inferior data around the world.  The consequence of this could be that although the U. N. Intergovernmental Panel on Climate Change (IPCC) has reported a rise of 0.64ºC (1.15ºF) in the planet’s average temperature in the last 50 years, the margin of error of that evaluation may be larger than the increase itself.  Third, weather stations sited in cities may record higher temperatures than warranted because of the known effects of asphalt pavement and absence of foliage on urban temperatures.  Additionally, sites that originally may have been isolated may have become urbanized as time passed, possibly altering the attributes of the sites.  Fourth, the major research groups analyzing temperature records selected stations preferentially having long-term data available, thus eliminating many other data sources that are available.  Fifth, methods and sites of measurement have changed or moved over the years, adding uncertainty to the results.  In Muller’s view, previous attempts to compensate for such variability may have introduced significant errors into the results.

BEST’s research methods.  In order to address these and other problems, BEST, directed by Muller, analyzed 1.5 billion discrete records from 15 preexisting temperature archives. Their objectives were to

1. create a single comprehensive raw data set in a common format for existing surface (land only) temperature data,
2. review existing data and methods of error analyses to assess their advantages and limitations,
3. develop improved and alternative methods of analysis to overcome some of the limitations identified, and
4. publish the new global surface temperature record, and its analysis, in order to  provide a basis for all parties to examine the data for further study.

The BEST project has concluded that global warming is real. The results are presented in four manuscripts, all available online (see Details below). There is also a summary available here.  The BEST study finds that “global warming is real”, having warmed by just under 1.0ºC (1.8ºF) since the mid 1950’s.  This conclusion should contribute significantly to eliminating doubts that global warming has been occurring since the 1950’s.  The reports address the major potential sources of error or bias raised by skeptics of global warming and identified by Muller, and show with high confidence that they do not affect the important conclusion reached.  This affirmation confirms the results of earlier studies, which skeptics had belittled because of their concerns over uncertainties in the data. 

Muller’s team made no effort address the question of whether the observed warming is due to human activity (mainly burning of fossil fuels to yield greenhouse gases), nor to predict future trends or effects on the planet.

Details

The first paper is titled “Berkeley Earth Temperature Averaging Process”.  The new statistical methods developed to analyze the merged data are described here.  They permit use of short, and discontinuous, records, permitting virtually all available data to be analyzed.  Such data had been ignored previously.  It develops a statistical weighting method that accounts for reliability of the data, and of their spatial distribution.  The method includes weighting only for spatial differences at fixed times, as well as “weather” effects, i.e., short term fluctuations due to local weather events.  In addition, a statistical “scalpel” is applied, with weighting, when discontinuities in the data arise.  The scalpel creates two distinct records from one original, broken at the point of a discontinuity. Such records can be included here, but would have been omitted in earlier analyses, because BEST developed methods of including short-term records.  This accommodates data of varying levels of quality without compromising the final result.  Methods for removing outliers, for assessing reliability of data points, and for weighting uncertainties were also developed for the BEST analysis.  The results are applied to develop a new global land temperature reconstruction from 1800 to the present with due accounting for identifiable sources of error, as summarized above.  Data from 7280 weather stations in the Global Historical Climatology Network (GHCN) were analyzed in the BEST assessment.

The overall result shows an increase in the earth’s long-term average temperature from the 1950’s to the present of 0.911 ± 0.042ºC (1.640 ± 0.076ºF; 95% confidence for statistical and spatial uncertainties; see the following graphic).  This result is stated to be consistent with the extent of global warming previously reported by the IPCC (0.64 ± 0.13ºC (1.15 ± 0.23ºF) from 1956 to 2005 at 95% confidence), but carries a considerably lower extent of uncertainty because of the new data included and the new methods of analysis. 

Result of the Berkeley Average Methodology applied to the GHCN monthly data from 1800 to 2010. Top plot shows a 12-month land-only moving average (red) and associated 95% uncertainty from statistical and spatial factors (grey with black edges). The lower plot shows a corresponding 10-year land-only moving average and 95% uncertainty. 
Source: BEST; http://www.berkeleyearth.org/

The following graphic shows the temperature differences, or “anomalies”, for all land areas of the planet, identified from the results of the BEST analysis, for two time intervals.  The color coded “anomalies” are based on the average temperatures in the earlier decade as the reference values. Warming is observed over all continents, with the greatest warming at high latitudes (northern regions of the map) and the least warming in southern South America.

Color-coded global projection maps showing the decadal average changes in temperature on the planet’s land areas, using the temperature scale on the right. In the upper plot, the comparison is drawn between the average temperature in 1900 to 1910, used as the reference, and the average temperature in 2000 to 2010. In the lower plot, the same comparison is made but using the interval 1960 to 1970 as the starting point.

           Source: BEST; http://www.berkeleyearth.org/

The time dependence of temperature changes arrayed across the land surfaces of the earth on a projection of the globe, from 1800 to the present, are also presented by BEST in this video.

In a second paper, “Influence of Urban Heating on the Global Temperature Land Average Using Rural Sites Identified from MODIS Classifications”, Muller et al. compared temperature records from urban settings to those from a selection of rural weather stations.  Their results confirm earlier analyses by others that urban warming has no biasing effect on assessing the extent of recent increases in global temperature.

The third paper, “Earth Atmospheric Land Surface Temperature and Station Quality in the United States”, assesses use of weather stations in the U. S. ranked by the National Oceanic and Atmospheric Administration to be “poor”, compared to those ranked “OK”.  The analysis finds no statistically significant difference in temperature trends between the two groups of stations.  This suggests that trends remain apparent even when individual stations have large uncertainties in their data.

The fourth paper, “Decadal Variations in the Global Atmospheric Land Temperatures”, examines correlations between variations, in the range of years to decades, in global average temperatures and pronounced ocean circulation trends.  In the past, many climate scientists have emphasized the role of the El Nino Southern Oscillation (ENSO) in global climate patterns.  The analysis in this paper from BEST finds the strongest correlations not with ENSO, but with the Atlantic Multidecadal Oscillation.  In addition, this paper separately includes an analysis of data from 30,964 world-wide weather stations with data available from 1950.  This number is much greater than had been used in earlier studies (the National Oceanographic and Atmospheric Administration, NASA Goddard Institute for Space Science, and a collaboration of the Hadley Centre of the UK Meteorological Office with the Climate Research Unit of East Anglia).  The new analysis conforms well to the earlier results.    

Conclusion.  Richard. A. Muller, a former skeptic concerning the validity of temperature trends that establish global warming, has conducted a rigorous analysis of all available land-based temperature records around the earth.  He and the BEST team developed new statistical methods to overcome several objections that he and others had raised earlier concerning historical temperature records.

Muller and the BEST team concluded that global warming is real.  This eliminates one argument that skeptics of global warming can use in denying the reality of the effect.

BEST did not at this time seek to evaluate whether the global warming they confirm is due to man-made effects.  The IPCC, however, has shown convincingly that man-made contributions to greenhouse gases in the atmosphere are required, in climate models, to reproduce the historical record of global changes in temperature.  The following graphic

Pale Blue shading: 5 to 95% range for 19 distinct simulation runs from five climate models using only natural processes that may affect global temperature.  Pink shading: 5 to 95% range for 58 distinct simulation runs from 14 climate models using both natural processes and man-made contributions (greenhouse gas emissions) that may affect global temperature.  Black line: actual temperature observations, averaged.

Source: IPCC ; http://www.ipcc.ch/publications_and_data/ar4/syr/en/figure-spm-4.html

shows specific continental, overall global, global land, and global ocean temperature simulations from climate models, compared to actual observations (black line).  The pale blue simulations omit man-made contributions to global warming, such as emissions of greenhouse gases.  It is seen that in none of the regions or global cases do the pale blue simulations reproduce the temperature observations beyond about 1950, but do satisfactorily trace the temperature data before 1950, before emissions became significant.  The pink simulations include man-mad contributions to global warming.  These simulations satisfactorily trace the observed temperature record over the entire time period from 1910 to 2005.  It may be concluded from these comparative simulations that man-made emissions of greenhouse gases are highly likely to be responsible for the increase in global temperature since about 1950.

© 2011 Henry Auer